The Power of 10 – a Recap
In the last article, I considered the impact of the players who have played the “Number 10” or Attacking Centre Midfield (AMC) role over the last 3 season (The Power of 10). Rogic is clearly the most effective goal scorer and held his own in defensive actions, whilst Johansen was the chances created king at 1.38 per 90 minutes. In the comments section, poster SK93 made an interesting point – “I think one factor that you can’t quantify but that is quite visibly apparent is that Rogic seems to make the entire team play better – general movement and attacking always seems more fluid/promising when he’s on the pitch.”. An excellent challenge to the numbers! Does Rogic make the team play better; is he the glue that makes Celtic a more effective attacking team?
The hypotheses of this piece, therefore, is “Celtic are a more effective attacking team with Rogic starting”.
Passing Networks
Before we get into that, a slight detour. I intend to consider attacking indicators such as passes complete, chances created and shot accuracy, as well as overall win/loss record. However, what would be excellent would be to see whether the overall attacking efficacy of the team is improved based on one players input. I will hold up my hand and say I do not have the data to show this but would draw your attention to those that can. Please read this link to see the work that can be done on passing networks given more budget, time, and software than I possess. I recommend you read this http://www.statsbombservices.com/explaining-xgchain-passing-networks.
All that being a case of “here is the speed boat you could have won”, what I can do is compare the creativity of the team with Rogic starting and not. What we can do is infer rather than “prove” beyond reasonable doubt.
The Rogic Effect
Under Deila, Rogic started around half the 2015/16 matches and did not start the other half. However, the games he started were almost exclusively non-European matches – he only started one match in the Europa League that season. He therefore did not start eleven European matches including the Champions League Qualifiers.
Under Rodgers, conversely, Rogic started nine Champions League matches including all the Group stage “hard” games. He missed three Champions League qualifiers as starter, and has missed the last thirteen domestic matches.
As we know, and I will show in a later article using numbers, Celtic play almost two different types of football – domestic fare compared to ties against some of the best teams on the plane, in Europe. The difference between Celtic in Europe and Celtic domestically is stark in terms of style of football and therefore performance statistics outcomes. Therefore, the “Rogic effect” is lost when you consider the opposition faced in both the two seasons I have data for. (NOTE – If anyone want to see some of the data I can provide).
Big Games
So, what can we do? I have again used the concept of “Big Games” to differentiate. A reminder of the definition of a” big game”:
- All matches against top 4 sides in the SPFL Premier League;
- Any domestic cup semi-final and final; and
- All European matches.
Looking at performances with and without Rogic in big games, the results and attacking team performances are better WITHOUT Rogic:
| P | W | D | L | F | A | Pts Average | |
| WITH ROGIC | 23 | 11 | 4 | 8 | 45 | 37 | 54% |
| WITHOUT ROGIC | 23 | 14 | 5 | 4 | 44 | 25 | 68% |
| Pass Complete | Pass Incomplete | Pass % | Chances Created | Chances Attempted | Chance Success | ||
| WITH ROGIC | 447 | 81 | 85% | 3.6 | 5.1 | 41% | |
| WITHOUT ROGIC | 451 | 87 | 84% | 4.6 | 6.8 | 40% | |
| Shot on target | Shot off target | Total Shots | Shot Accuracy | On Target Conversion | All Shot Conversion | ||
| WITH ROGIC | 5.7 | 10.2 | 15.9 | 36% | 34% | 12% | |
| WITHOUT ROGIC | 6 | 10 | 16 | 38% | 32% | 12% |
There are some odd outcomes here. Despite having better results without Rogic, the shots created and the accuracy of them is virtually the same – there may be something here on shot and chance quality but I do not have the Expected Goals data for last season.
The games WITH Rogic take in all the Champions League groups stage matches versus a far higher standard of opposition than the previous year’s Europa League, most of which Rogic missed. At the risk of appearing to dismiss the data because it does not fit the narrative, I am minded to consider that the quality of the opposition (much higher WITH Rogic playing) is the key factor here.
Domestic Games (not “Big Games”)
Finally, let’s consider relative performances when playing in run of the mill domestic games – i.e not “big games” as defined above.
| P | W | D | L | F | A | Pts Average | |
| WITH ROGIC | 26 | 20 | 5 | 1 | 72 | 19 | 83% |
| WITHOUT ROGIC | 18 | 16 | 2 | 0 | 50 | 5 | 92% |
| Pass Complete | Pass Incomplete | Pass % | Chances Created | Chances Attempted | Chance Success | ||
| WITH ROGIC | 548 | 83 | 87% | 6.0 | 7.0 | 46% | |
| WITHOUT ROGIC | 512 | 87 | 85% | 5.1 | 7.3 | 41% | |
| Shot on target | Shot off target | Total Shots | Shot Accuracy | On Target Conversion | All Shot Conversion | ||
| WITH ROGIC | 9 | 15.3 | 24.3 | 37% | 31% | 11% | |
| WITHOUT ROGIC | 7.9 | 12.4 | 20.3 | 39% | 36% | 14% |
Again, Celtic win more WITHOUT Rogic than with, mainly due to him missing a large part of the Rodgers (r)evolution where Celtic are compiling a perhaps unprecedented run of domestic success. However, perhaps counter intuitively, Celtic pass considerably more WITH Rogic in the team (548 to 512) and create more chances (13 to 12.4) and have more shots (24.3 to 20.3). This does not prove that this is down to Rogic but indicates that at least in games against the lesser opposition, when he plays, there is some increase in attacking output.
Summary
There is evidence that in the “bread and butter” domestic games having Rogic start means the team achieve more cohesion evidenced by more successful passing, and create more chances and shots. However, based on data from “big games”, the opposition may be a more important factor than one player in this instance. Looking across the results, it is not compelling that having Rogic in the team materially impacts the overall performance. Subjectively, I still prefer to watch Celtic play when Rogic is on the park!
To borrow a phrase much loved by the SFA in selective disciplinary cases, the original hypotheses is “not proven”. Perhaps a more interesting theme which I will return to, is that under Rodgers, Celtic do not appear to create more chances or shots on goal when compared to the Deila reign.